verification-before-completion

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the agent to run local shell commands such as cargo test, npm run build, and ripgrep to provide evidence of task completion and code correctness.
  • [COMMAND_EXECUTION]: Delegates the execution of potentially verbose or resource-intensive tasks, such as running test suites and pre-commit hooks, to an external hyperpowers:test-runner agent.
  • [PROMPT_INJECTION]: Employs rigid process instructions like 'Iron Law' and 'LOW FREEDOM' to ensure the agent does not shortcut verification steps.
  • [PROMPT_INJECTION]: Exhibits an indirect prompt injection surface by processing potentially untrusted data from command outputs and task management tools. 1. Ingestion points: Command stdout/stderr and bd task success criteria. 2. Boundary markers: None present. 3. Capability inventory: Local command execution and delegation to external agents. 4. Sanitization: No sanitization or validation of external content is performed.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 05:19 PM